| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 79 No. 2 February 1998, pp. 537-554
Copyright ©1998 by the American Physiological Society
Northwestern University Medical School, Department of Physiology, Chicago, Illinois 60611-3008
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Mason, Carolyn R., Lee E. Miller, James F. Baker, and James C. Houk. Organization of reaching and grasping movements in the primate cerebellar nuclei as revealed by focal muscimol inactivations. J. Neurophysiol. 79: 537-554, 1998. Two monkeys were trained to point to targets and to retrieve fruit bits from a Kluver board, bottles, and tubes. Once proficient in the tasks, the macaques underwent aseptic surgical implantation of a recording chamber over the cerebellar nuclei on the side of their preferred hand. After recovery from surgery, a series of mapping penetrations were completed to identify task-related areas within the cerebellar nuclei. Muscimol (4- 16 µg; 1-2 µg/µl) was pressure injected at different sites within the forelimb zone, and the resultant deficits were observed as the monkeys performed the behavioral tasks. Quantitative measures of task performance were supplemented by direct observation of live and videotaped performance. The locations of nuclear inactivation sites were reconstructed from marking lesions and tracks visible in histological sections. Injections placed in the cerebellar interpositus nucleus and adjacent regions of dentate caused a variety of deficits in forelimb function. A prominent anteroposterior specialization was apparent within the forelimb zone of this intermediate nuclear region. Injections into the anterior interpositus nucleus and adjacent dentate impaired preshaping of the hand and the manipulation of objects, whereas injections placed more posteriorly in posterior interpositus nucleus and adjacent dentate produced deficits in the aiming of reach and the stability of the arm. During anterior injections, the monkeys failed to adequately extend their fingers in preparation for target contact, as documented for >85% of the reaches in the pointing task of monkey J. Up to 38% of the fruit bits it attempted to retrieve from the Kluver board were dropped. In comparison, during posterior inactivations, 15% were dropped and during control conditions 3% were dropped. The monkeys made significantly greater pointing errors during posterior inactivations (11 times for monkey J and 4 times for monkey C) than during anterior inactivations (8 times for monkey J and 2 times for monkey C). We refer to the region producing hand deficits as the anterior hand zone and the region producing reaching deficits as the posterior reach zone. These results are discussed in relation to the problem of achieving spatiotemporal coordination in the large population of nuclear cells that participate in any given movement. The results do not favor the hypothesis that coordination is achieved through a selection of Purkinje cells along beams of parallel fibers. Instead, it is proposed that distal and proximal musculature is coordinated by the adaptive influences of climbing fiber input to Purkinje cells. We envision a relatively nonspecific recruitment of anterior and posterior nuclear cells due to positive feedback in the limb premotor network, which then is shaped into an appropriate spatiotemporal pattern of discharge through the inhibitory input from Purkinje cells.
Experimental evidence from a variety of techniques has established a mediolateral organization of the primate cerebellum, dividing the cerebellar cortex and the underlying deep nuclei into three functional zones. The most medial zone consists of the fastigial nucleus and the overlying vermal cortex. This zone projects to the medial descending system, which mainly is involved in the control of posture and locomotion (Kuypers 1981 Behavioral task description
Two male macaques were trained in a variety of behavioral tasks. The first animal, monkey C, was a Macaca fasicularis weighing 5.5 kg. The second, monkey J, was a M. mulatta weighing 7.5 kg.
POINTING TASK.
The equipment for the pointing task was a touch pad and target array. The touch pad was a capacitive proximity detector that was positioned by the monkey's side at waist level. The monkey's hand rested easily on it out of view. The target array consisted of a 6 × 6 grid of capacitive proximity detectors, each 2.5 × 2.5 cm. Embedded in each of the central 16 touch detectors was a green target light emitting diode (LED). This allowed reaching errors to be detected in all directions around a given LED (Fig. 1A). The target array was oriented vertically and positioned in front of the monkey at a comfortable reaching distance (Fig. 1B).
BOTTLE TASK.
Various bottles were used to assess the monkey's ability to aim correctly for the bottle and shape its hand properly for insertion into the bottle. The successful removal of the fruit bit from inside the bottle also gave an indication of how well the monkey was manipulating small objects. All the bottles were opaque except for the largest one. Opening sizes were 2, 3.2, 3.7, and 6.8 cm. The smaller openings required better aim and hand shaping than did the larger openings. The bottles typically were offered to the monkeys at a comfortable reaching distance in front of them. At times, bottles were held on the fringes of their work space.
TUBE TASK.
Monkey C failed to retrieve fruit bits from the bottles during some inactivations, but the opaqueness of the bottles prevented observing the suspected impairment of hand function. Consequently, the tube task was introduced with monkey J to enable the investigator to observe the monkey's hand function when the monkey's view of its hand was obstructed. Each tube had a vertical face plate with an opening to the tube (Fig. 1D). The transparent tube extended away from the monkey perpendicular to the face plate. An opaque sliding door was positioned in slots 1.2 cm in front of the tube opening. The door was opened manually by the investigator sliding it laterally. An infrared diode and phototransistor detector were located above and below the opening between the door and the tube. The detector picked up three levels of signal. The lowest level indicated the door was closed. An intermediate level indicated the open door, and the highest level indicated when the sensor was occluded by the monkey's limb. The monkey's performance could be timed by the level changes. After a random interval of contact with the touch pad by the monkey, the door was opened and the monkey was given the opportunity to retrieve the fruit bit. If successful, the monkey placed the fruit bit in its mouth before returning to the touch pad (Fig. 1F).
KLUVER BOARD.
Another assessment tool of reaching and hand deficits was the Kluver board. The Kluver board had nine slots arranged in three parallel rows of three. Each slot was 1 cm wide, 4 cm long, and 1.2 cm deep. A small piece of fruit was placed into each slot. The board typically was presented with the slots oriented either vertically or horizontally. The angle of the board in relation to the monkey and its distance from the monkey varied.
Surgery
After training, the monkeys underwent surgery for implantation of a recording chamber and head holder. Under halothane anesthesia and aseptic conditions, a stainless steel recording chamber was placed over the parietal lobe ipsilateral to the trained limb in stereotaxic alignment with the deep cerebellar nuclei, ~8 mm posterior and 3 or 4 mm lateral. The chamber of first monkey was vertical in the sagittal and frontal planes. The angle of penetration of the second monkey slanted 5° anteriorly and 5° medially. All hardware was anchored with bone screws and dental cement. Both monkeys were treated with the prophylactic antibiotic cefazolin (Keflex 35 mg/kg im) before and after surgery. Monkeys received buprenorphine hydrochloride (Buprenex; 0.01 mg/kg im) for pain control for 2-3 days after surgery. For all behavioral tasks, the monkeys sat in a primate chair with their heads fixed.
Electrophysiological recordings
A series of mapping penetrations was made to identify the subdivisions of the cerebellar nuclei on the basis of their single and multiunit activity. This tentative identification was histologically confirmed later after the animal had been sacrificed. We recorded neural activity with epoxy-coated tungsten microelectrodes mounted in a modified Narashige microdrive while the monkey performed the pointing task described earlier. Electrode penetrations passed through parietal cortex and into the cerebellar cortex, which was identified on the basis of the characteristic Purkinje cell and granular layer activity (van Kan et al. 1993a
Muscimol inactivations
Several weeks to months after the mapping penetrations, muscimol was injected at various nuclear locations through a cannula mounted on the microdrive. The doses and volumes are listed in Table 1. For monkey C, doses were typically 4 µg (range 2-4 µg) diluted in volumes of 1-2 µl. For monkey J, the doses ranged from 4 to 12 µg diluted in 2 µl except for the largest dose, which inadvertently involved 6 µl. If no effect was observed after ~30 min, a second injection was given in some cases either at the same or a more ventral site. In Table 1, time of deficit onset is given from the time of the first injection. In the final inactivations (J9.5, J10, and J10.5), the cannula had an indwelling fine wire electrode that allowed the monitoring of multiunit activity, which disappeared after injection of the muscimol.
Data collection
Various signals related to the monkey's task performance, were collected on PDP 11/73 and Macintosh IIcx computers. If a single unit had been well isolated, output pulses from the discriminator were low-pass filtered (10-ms time constant) to provide a signal proportional to the unit's firing rate. The pointing task-related signals included touch pad contact, target LEDs, and target array contact. During the performance of the tube task, the phototransistor detector signal was also collected. The same behavioral data were collected during control and inactivation sessions, and performance also was videotaped to allow direct assessment of the movements and any skill deficits. Control data for monkey C was collected immediately preceding each inactivation. Instead, control data for monkey J was collected on other days when inactivations were not attempted. For various reasons, the monkeys did not necessarily perform all of the tasks during each inactivation.
Analyses
Our first analyses focused on measures of goal achievement, for example, the accuracy of pointing or the success at removing fruit from the tube. Often, during inactivations, the monkeys could complete the task in a manner that was rewarded even though they employed different strategies than under control conditions. In these cases, the measures of goal achievement did not reflect the impairments the monkeys exhibited or the compensations the monkeys used to accomplish the goals. Other measures were introduced to examine the changed strategies and to determine if there was a site dependence of the different impairments observed during the inactivations. These measures focused on how the monkey performed the tasks. Analyses used both computer and videotaped records of monkey behavior. When possible, the Mann-Whitney U test was used to compute the significance of differences between conditions.
POINTING TASK.
A measure of a goal achievement was the accuracy of the monkey's pointing. For each reach, a distance error was calculated by subtracting the initial row and column touched from the target row and column. The resulting two numbers were squared and then summed. The square root of that sum was defined as the distance error for that reach. Two other measures were implemented to document the manner in which the monkeys accomplished the pointing. One was the number of different rows the monkey touched as well as the number of rows touched simultaneously during a given trial. The other was the sequence of contacts the monkey made on the target array. Problems with hand stability or shape resulted in an atypical patterns of contact. A count of the number of rows touched gave an indication of how steady the monkey could hold its limb during the point. The number of rows contacted simultaneously gave an indication of the monkey's hand shape. If the monkey was pointing with its fingertips, it would touch fewer rows simultaneously than if its fingers were flexed either at the initiation of contact or as the contact continued. Another way to address stability and hand shaping was to look at the sequence of contacts the monkey made during its contact with the target array. The contact sequence was determined by the rows touched sequentially from the monkey's first contact with target array until the monkey removed its hand for the return to the touch pad. The contact sequence indicated if the monkey was touching with its finger tips and sliding down the surface of the target array or if the hand was making contact with multiple adjacent or nonadjacent rows at once indicating the hand was flattening against the surface of the target array.
BOTTLE TASK.
Monkey C had the two smaller bottles and sometimes the largest bottle presented to it in each inactivation. Monkey J was presented the three larger bottles. From the video tapes, the number of attempts required by each monkey to insert its hand into the bottles was counted. The first approach was counted as one and each retreat and reapproach incremented the count. The count continued until the monkey successfully inserted its hand into the bottle or gave up.
TUBE TASK.
The transparent tube allowed videotaping the hand movements, and the detector in the entryway allowed timing of hand insertion and removal. If the monkey were having a problem with aim, we might expect a prolongation of the time needed for the reaching component of the task but not necessarily of the time spent in the tube picking up the fruit. The prolongation of either time period should be accompanied by videotaped documentation of deficits. Measures of the monkey's performance could be made from the videotape.
KLUVER BOARD.
The percentage of fruit bits dropped while performing the Kluver board was calculated for monkey J by counting the number of dropped fruit bits when the slots were in the vertical position and dividing by the number of slots (range 8-29) from which the monkey had the opportunity to remove fruit bits. This calculation was not computed for monkey C because the Kluver board was not presented in a consistent manner.
Histology and reconstruction
Marking lesions were made after completion of all the inactivations several days before the animals were perfused. In addition, electrolytic marking lesions were made after data collection in inactivations J9.5, J10, and J10.5 using the fine wire electrode in the cannula.
Muscimol injections were placed at discrete cerebellar nuclear sites in a forelimb region that had been identified by microelectrode mapping (cf. METHODS). The variety of behavioral deficits that resulted were classified as predominant problems in forelimb reaching, hand use or, in one case, use of the hindlimb (Table 1). Despite severe impairment of reaching or hand motor control that was apparent to the investigator and in the video tapes, both monkeys often could complete the behavioral tasks with only modest degradation of performance. The monkeys could make adjustments in their movements that minimized the effects of their deficits on task performance. For example, the hand could grip an object to stabilize a poorly controlled arm, or repeated reaches by a well-controlled arm could compensate for an imprecisely shaped hand. Careful observation of live and videotaped task behavior was required to establish the exact nature of the disability produced at a particular inactivation site. After briefly categorizing the observed deficits, results are presented in the context of the four tasks.
Pointing task
The pointing task included two components, reaching and hand use. In the course of the mapping recordings (METHODS), we encountered single units in the deep nuclei having discharge patterns that were related to the transport phase of the reach (Fig. 1A) or to the contact with the target array (Fig. 1B). Injections of muscimol into the nuclei caused performance deficits which also could be related to one of these two components of the pointing task. Muscimol-induced deficits during the reaching component included both decreased accuracy and instability of the limb. Less obvious, perhaps because the task did not include manipulation of an object, were deficits in the hand-use component. These mainly resulted from abnormal hand shape and digit extension as the monkey attempted to touch the lighted target square.
Bottle task
The bottle task included two components, reaching and properly preshaping of the monkey's hand for insertion into the bottle. During control conditions, the monkeys could insert their hands on the first attempt as is shown for monkey C in Fig. 5A. During inactivation C10b, monkey C had its hand properly shaped for entry into the bottle but required many attempts (Fig. 5B). The tracings in Fig. 5D depict three such successive attempts. In the first and third attempt, the monkey's hand went over the bottle. On its second attempt, the hand went off to the left of the bottle with the fingers curling as the hand passed by the bottle opening. Monkey J exhibited similar aiming problems during inactivations J9.5, J10, and J10.5. Monkey J also exhibited stability problems. During inactivations J9.5, J10, and J10.5, if the monkey managed to get its hand on the side of the bottle, it would try to pull its hand back to the opening. In most instances, the hand would fall off the bottle and tremor would ensue.
Tube task
The tube task stressed aim but also required proper hand preshaping for entry into the tube and good stability and manipulation skills once in the tube. Under control conditions, monkey J was observed, both live and on videotape, retrieving the fruit bits in very stereotypical manners for each tube used. For example, during control trials with the smallest tube, the monkey retrieved the fruit bit by flexing its fingers to slide the fruit toward the end of the tube where it would catch it between the finger pads and dorsum of the thumb, dropping <5% of the fruit. If using the tube with the shelf, the monkey removed the fruit from the shelf using a pinch grasp successfully 73% of the time. It always retrieved fruit that had dropped to the bottom of the tube. Figure 6A graphically presents the reaching and manipulation times during one control session with the tube with a shelf. The mean reach time (0.23 ± 0.05 s,
Kluver board
The Kluver board task has three phases. During the first phase, the transport phase, the monkey must aim for the slots as well as prepare its hand for the removal of the fruit bit. During the second or removal phase, the monkey must hold its hand steady and use its thumb and forefinger in a coordinated manner to remove the fruit bit from the slot. The remaining lateral fingers should be held quiescent out of the way. In the third phase, the monkey must maintain its grip of the fruit bit until it brings its hand to the mouth and then let go. Under control conditions, both monkeys accomplished all three phases skillfully. During inactivations, the monkeys' performance of the Kluver board task could be hindered by reaching or hand deficits.
Correlation of deficits with inactivation sites
The reconstructed locations of the nuclear inactivation sites in monkey J are illustrated on a set of tracings from frontal sections through the dentate and interpositus nuclei in Fig. 7. Each inactivation site produced a distinctive set of deficits as judged by the quantitative measures of task performance summarized in Tables 2-4 along with careful observation of live and videotaped movements. The inactivations located at the more anterior sites caused different types of hand deficits, marked in Fig. 7 (×), whereas inactivations at more posterior sites caused deficits in reaching (
A variety of deficits in forelimb function resulted when the GABA agonist muscimol was injected at discrete sites in the cerebellar interpositus nucleus and adjacent regions of dentate. The impairments produced by 15 inactivations in two monkeys revealed a prominent anteroposterior specialization of function within the forelimb region of the intermediate cerebellum (Fig. 8). The more anterior injections, into caudal NIA and adjacent dentate, produced various deficits in hand shaping and manipulation, whereas injections more posteriorly in NIP and adjacent dentate impaired the aiming of reach and the stability of the arm. We refer to the region producing hand deficits as the anterior hand zone (AHZ) and the region producing reaching deficits as the posterior reach zone (PRZ).
Skill deficits and compensations
The muscimol injections into the cerebellar nuclei described in this paper resulted in profound deficits, although this was not always apparent from an analysis of task success measures. Both monkeys could, to a large degree, complete the tasks required of them in spite of the deficits caused by muscimol inactivations. The deficits became most apparent when the strategies the monkeys used to accomplish the tasks were fully examined. Other investigators (Alstermark et al. 1981 Comparison with previous lesion studies
Large irreversible lesions of the intermediate and lateral cerebellum in monkeys generally resulted in the full constellation of reaching and hand deficits seen in the present study (Botterell and Fulton 1938 Coordination of forelimb movements
It is accepted widely that coordination of movement is one of the main functions of the cerebellum (Ito 1984
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
). Monkeys with their fastigial nuclei removed (Carrea and Mettler 1947) or inactivated (Thach et al. 1992b) have difficulty sitting unsupported, standing, or walking.
; van Kan et al. 1993b), suggesting that they be included as part of the intermediate zone. Like NIA and NIP, these regions of dentate project to the magnocellular red nucleus (Kennedy et al. 1986) and to the primary motor cortex by way of ventrolateral thalamus (Middleton and Strick 1997). Through these projections to rubrospinal and corticospinal pathways, the intermediate zone is poised to control the independent use of the limb in reaching and grasping (Kuypers 1981). Lesions centered on the intermediate zone in monkeys result in impaired reaching and poor use of the digits (Growdon et al. 1967; Mackel 1987).
; Middleton and Strick 1997; Rispal-Padel 1993). This major part of dentate can be considered the nuclear component of the lateral zone of the cerebellum, leaving dorsomedial dentate as part of the intermediate nuclear zone.
; Thach et al. 1992a,b). According to Thach's analysis of these data, each of the cerebellar nuclei contains a complete map of body parts with the leg located anteriorly, the arm at intermediate sites, and the head located posteriorly. Thach and colleagues further proposed that the myotomes within a given body part are represented along mediolateral gradients in the cerebellum, which could allow parallel fibers to coordinate muscle synergies about proximal and distal joints.
) confirmed the presence of a leg representation anterior and an arm representation posterior within NIA and adjacent parts of dentate. Within NIP and adjacent dentate, many additional arm-related units were found mixed with a few leg-related cells, and head-related units were found in the most posterior regions of NIP. In the forelimb regions of both NIA and NIP, the most intense unit activity was noted during multijoint movements as opposed to single-joint motion (van Kan et al. 1993b). Individual neurons were not specifically related to motion about individual joints (Thach et. al. 1993; van Kan et al. 1993b), suggesting that a finer representation of the forelimb may not be present in the intermediate nuclear zone. Subsequent studies stressed the importance of including a component of hand use for obtaining high-frequency discharge in both NIA and NIP (Gibson et al. 1996; van Kan et al. 1994).
, 1983). Within the arm zone, stimulation in the center of NIP produced shoulder movement, whereas stimulation of ventral NIA and dorsal NIP produced different kinds of finger and hand movement.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
View larger version (31K):
[in a new window]
FIG. 1.
Pointing task and tube task equipment and paradigm. A: front view of the target array showing the 2.5-cm square capacitive touch detectors. Target light emitting diodes (LEDs) are centered in the central 4 × 4 array of touch detectors. B: side view cartoon of equipment setup with a stick figure of the monkey with its hand resting on a touch pad on its lap plate and the target array positioned parallel to the monkey's frontal plane. C: this schematic of the reaching paradigm indicates that the trial begins with the monkey resting its hand on the touch pad for a variable interval. A target LED then lights. Monkey picks up its hand and reaches for the target. After contact with the target array, the monkey returns its hand to its touch pad to initiate the next trial. D: front view of the face plate of the tube. E: side view of the tube showing the location of the door and of the shelf. F: this schematic of the tube task shows that the monkey starts with contact with its touch pad before the investigator would open the door to the tube. Once the door opened, the monkey would initiate the reach for and enter the tube to retrieve the fruit bit placed previously on the shelf. Monkey would bring the fruit bit to its mouth before returning its hand to the touch pad to initiate the next trial.
). Upon exiting the cerebellar cortex, the electrode would pass through 1-2 mm of relatively quiet white matter with only an occasional fiber potential. Penetration of the nucleus would result in a sudden increase in activity. Single units had burst firing rates that could reach 200-400 spikes/s and spike amplitudes 
1 mV. The relation between neural activity and the monkey's behavior was determined, and, if an isolated unit was well related to the tasks, one or more data files were recorded.
), the interpositus nucleus and adjacent regions of dentate were recognized readily by the presence of neural activity related to voluntary movements of the forelimb. In addition, we found that recordings from posterior sites typically evoked activity during the transport phase of the pointing task as illustrated in Fig. 2A, whereas anterior recordings were better related to contact with the target array (Fig. 2B) and other activities that used the hand (illustrated later). These recording were an invaluable aid to the placement of muscimol injections.
View larger version (28K):
[in a new window]
FIG. 2.
Two nuclear neurons recorded while monkey J performed the pointing task. A: this tracing is an ensemble average of a nuclear neuron's activity during 15 reaches. Average is centered on target contact. Increase in activity began as the monkey lifted its hand off and touch pad and ceases abruptly before contact with the target array. This unit was recorded in the vicinity of inactivation J10.5. B: this ensemble average of neuronal activity during 14 reaches, recorded in the vicinity of J6.5b, had a brief burst of activity as the monkey removed its hand from the touch pad then shut off during the transport phase of the reach. A second more prominent burst occurred just after contact with the target array. A smaller third burst occurred as the monkey's hand approached its touch pad.
View this table:
TABLE 1.
Summary of inactivations in monkeys C and J
. An injection of 1 µg of radio-labeled muscimol in 1 µl saline resulted in a radioactivity profile that dropped off steeply to background levels ~1 mm from the injection site, when measured 10 min postinjection. At later times, the peak concentration was diminished appreciably, whereas the spread was only slightly expanded, presumably due to strong receptor binding and powerful uptake mechanisms for this long-acting 
-aminobutyric acid (GABA) agonist. The spread of hypometabolism as measured in the cerebral cortex was considerably greater; this was attributed to neuronal interaction in local circuits rather than to drug diffusion.
measured dye diffusion and found that a 2-µl volume spread 2.4 mm from the injection site, whereas 4 µl spread up to 3.7 mm. Although the strong binding of muscimol should limit diffusion to shorter distances, our larger injections should nevertheless spread a few millimeters, which would explain why the 6-µl volume inadvertently given in inactivation J7.5b inactivated both anterior and posterior portions of the forelimb zone, thus explaining the mixed hand and reach deficits. However, the 2 µl used for most injections probably were much more confined, which accounts for the very different hand deficits observed with inactivations that were only 1 mm apart in the anteroposterior dimension. The unique hand deficits observed at J5.5, J6.5, and J7.5 suggest that the 2-µl injections of muscimol remained functionally in a very localized region.
). The 3-D model then could be viewed from different angles with different degrees of transparency. Figure 8 is a horizontal view of this model illustrating the A-P and M-L coordinates of the cerebellar nuclei. The injection sites for monkey J, a left-handed animal, are shown on the left.
View larger version (72K):
[in a new window]
FIG. 8.
Three-dimensional model of monkey J's cerebellar nuclei. Inactivation sites of monkey J are located on the left side of this reconstruction. Inactivation sites of monkey C are approximately located on the right side of the reconstruction (METHODS). ×, sites with hand deficits; 
, sites of reaching deficits; 
, mixed hand and reaching deficit; 
, hindlimb deficit.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
View larger version (34K):
[in a new window]
FIG. 3.
Monkey J's accuracy and contact sequence during control condition and inactivation J10.5. A: this graph shows the distance error during a control session expressed as percentage of reaches. In 95 % of the reaches, monkey J was on target (error = 0). In 5% of the trials, monkey J touched both the correct square and the square immediately adjacent simultaneously (error = 0.5). B: during inactivation J10.5, only 53% of the reaches were on target, 39% hit both the correct square and the adjacent square, and 8% of the reaches were off an entire square. N, number of reaches in A and B. C: traces taken from a series of digitized video frames during a reach in inactivation J10.5. D and E: representations of the contact sequence of monkey J during a reach to a target in row 2. D: monkey made an initial contact in row 2 followed by row 3. Sequence was reversed as the monkey removed its fingers from the target array. Monkey touched 2 rows simultaneously and a total of 2 rows during this reach. E: during inactivation J10.5, the monkey first made contact in row 2. Its fingers then slid up touch row 1 before slidding down to rows 3 and 4 before removing its fingers from the target array. Sequence of rows document the slide or sweep of the fingers down the surface of the target array. Monkey touched no more than 2 rows simultaneously during this reach and touched a total of 4 rows. F: this graph compares the number of rows the monkey contacted simultaneously and the total rows touched in each reach during a control session, during inactivation J10.5 and during J6.5b. During inactivation J10.5, the monkey touched significantly more rows than during the control session. During inactivation J6.5b, the monkey touched significantly more rows simultaneously as well as more total rows.
View this table:
TABLE 2.
Distance errors and atypical metacarpophalangeal and finger extension during pointing task performance by monkey J
View larger version (64K):
[in a new window]
FIG. 4.
Monkey J's hand position during control condition and inactivation J6.5b and the contact sequence and accuracy during inactivation J6.5b. A and B: digitized frames from videotapes. A: a frame taken at the moment of contact with the target array during a control session. Fingers are extended. B: during inactivation J6.5b, the monkey lacked complete finger extension at the moment of contact with the target array. Monkey made no effort to correct its hand shape. Instead the hand further flattened during the reach before removal from the target array. C: this panel depicts a typical reach to a target in row 2 during inactivation J6.5b. Initial contact is in row 2. As the hand continues to flatten on the target array surface, additional rows are contacted. Monkey touched 3 rows simultaneously and touched 4 rows in total. Means for this inactivation are shown in Fig. 3F. D: distance errors for the initial contact during inactivation J6.5b suggest that the monkey's aim was not greatly impaired. Hand shape was the problem.
View larger version (17K):
[in a new window]
FIG. 5.
Number of attempts to enter the bottle with a 3.2 cm opening during control conditions, inactivation C10b and inactivation C8b. A: under control conditions, monkey C inserted its hand into the bottle on the first try the 4 times it was presented the bottle. B: during inactivation C10b, monkey C's aim was impaired hindering successful insertion of its hand into the bottle. Only completed presentations are depicted. C: during inactivation C8b, monkey C had little difficulty aiming its hand properly. A finger caught on the rim of the bottle in the 4th presentation requiring the second attempt to insert its hand. D: traces taken from digitized video of 3 successive attempts by monkey C to insert its hand into the bottle during inactivation C10b. In D, 1 and 3, its right hand went over the top of the bottle opening. In D2, its hand went off the left side of the bottle. Monkey curled its fingers after passing the opening.
) and the mean manipulation time (1.18 ± 0.38 s, 
) are close to the average of all the control sessions for this tube (reach time 0.26 ± 0.03 s; manipulation time 1.23 ± 0.62 s).
View larger version (38K):
[in a new window]
FIG. 6.
Graphic representation of reaching and manipulation times during the tube task and single unit recordings. A: under control condition depicted, monkey J took an average of 0.23 ± 0.05 s to lift its hand off its touch pad and reach the tube opening (reaching time, D, and left vertical axis) and an average of 1.18 ± 0.38 s to retrieve the fruit bit from the shelf and exit the tube (manipulation time, , and right vertical axis). C: reaching times (mean 0.29 ± 0.04 s) and manipulation times (mean 3.73 ± 1.58 s) during inactivation J6.5b. B and D: recording of a single unit in the vicinity of inactivation J6.5b. B is the ensemble average of single-unit activity while the monkey performed the pointing task. Activity is centered on contact with the target array. Unit burst just before target contact. D is a single trace of the same unit during the tube task. Activity of the unit increases just before the removal of its hand from its touch pad. Activity continues during the reach and is maintained while the hand is in the tube and the fruit brought to the monkey's mouth. E: reaching times (mean 0.34 ± 0.14 s) and manipulation times (mean 2.07 ± 1.01 s) during inactivation J10.5. F: single trace of activity during the tube task showing a burst of activity during the reach and return to baseline levels during manipulation time. This unit was shown during the pointing task in Fig. 2A and was in the vicinity of inactivation J10.5.
View this table:
TABLE 4.
Fruit bits dropped by monkey J during performance of Kluver board task
). The filled circle indicates a site of mixed hand and reaching deficits and the filled square is the site of a marking lesion.
View larger version (19K):
[in a new window]
FIG. 7.
Tracings of interpositus and dentate nuclei and inactivation sites for monkey J. Each tracing was made from a histological section at equal intervals. ×, sites of hand deficits; , site mixed hand and reaching deficits (J7.5b); , sites of reaching deficits; 
, marking lesion.
View this table:
TABLE 3.
Distance errors during pointing task performance by monkey C
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Armand and Kalby 1992) have noted the need to look beyond goal accomplishment to the skills with which the tasks were performed. During inactivations, some of the stereotypical movement patterns habitually used by our monkeys were replaced by less skilled movements combined with compensatory strategies that allowed the achievement of the task goal despite poor motor control.
; Growdon et al. 1967; Gilman et al. 1976). Electrolytic lesions destroying just the interpositus and dentate nuclei produced a similar constellation of deficits (Goldberger and Growdon 1973). For cases in which either dentate or interpositus were destroyed individually, the main finding was considerably less impairment of limb stability as compared with the combined lesions (Goldberger and Growdon 1973). Several of the injection sites in the present study were located in the white matter between the two nuclei and undoubtedly inactivated cells in both interpositus and adjacent regions of dentate, in zones that are known to project to magnocellular red nucleus (Kennedy et al. 1986) and to primary motor cortex (Middleton and Strick 1997). Although separated by white matter, these adjacent regions of interpositus and dentate are functionally related, and we consider both as contributing to the intermediate nuclear zone controlling the independent use of the limb in reaching and grasping (INTRODUCTION).
reported that the feline NIA projects densely throughout the magnocellular red nucleus, whereas NIP terminals are confined to a shell surrounding the core of the nucleus. A similar pattern in monkey is apparent from a comparison of the anterior and posterior cases presented by Kalil (1981) in her study of cerebellar projections to thalamus. This distinction may relate to the microstimulation and sensory field maps reported by Larsen and Yumiya (1980), who suggested a distal limb representation in the core of red nucleus surrounded by shells of proximal representation. The AHZ may influence hand movements through its projections to the core of red nucleus and PRZ may influence reaching through its projections to the shell. AHZ terminals also are found in the shell regions innervated by PRZ, which may help to coordinate hand and reach function. The projections via ventral thalamus to motor cortex also would be expected to differentially affect proximal and distal musculature, but the intricacy of thalamic organization complicates the analysis of these relations.
to produce irreversible lesions at several nuclear sites, some of which explored the anteroposterior dimension. The lesions were relatively large, and the histological reconstructions suggest that both the AHZ and PRZ were damaged in most of the cases. Such cases resulted in both reach and hand deficits. However, one lesion, reported as being restricted to the posterior third of the deep nuclei, did not impair voluntary limb movement at all; rather, the monkey exhibited a head tilt and postural ataxia. We observed a head tilt only in inactivations that were located histologically in the ansiform or paramedian lobules of the cerebellar cortex. Our procedure was to make a single, discrete microinjection, and these posterior sites were too distant to consider spread of inactivation to the PRZ (METHODS). Instead, Mackel's procedure was to make multiple injections spaced over a few millimeters, so it is possible that his posterior case damaged ansiform and/or paramedian lobules.
; Uno et al. 1973) used a cooling probe to study how reversible inactivation affects movement about the elbow in a tracking task. Cooling of the lateral dentate resulted in hypermetric movements or the overshooting of the target, whereas cooling of the lateral interpositus, both anterior and posterior and also the medial dentate, resulted in hypometric movements. Both the hypermetric and the hypometric movements should result in aiming deficits in a less restrained task. Hand movements were not tested. In baboons trained to perform a pointing task, Beaubaton and Trouche (1982) noted impaired aiming with an increased dispersion of the movements during dentate cooling. The probe placement and the isotherms from the cooling appear to cover all of the dentate except for the medial anterior and most caudal parts. These results are consistent with our finding of impaired aim during muscimol inactivation of PRZ. Their isotherms suggest a sparing of AHZ, and they did not report any hand deficits.
), three inactivation sites in posterior dentate are shown, each of which resulted in reach deficits. The most medial of these is within PRZ, where we consistently observed reach deficits. The most lateral site at this posterior location may be in the zone identified by Middleton and Strick (1997) as projecting to the ventral premotor area; units here are selectively involved in visually cued arm movement. As noted earlier, visual-triggering specificity is a summary feature of Thach's dentate unit sample (Thach et al. 1992a,b). It would be helpful to know the anteroposterior coordinates of the dentate sites producing hand deficits (Thach et al. 1992b) to determine if they were within the AHZ.
). We were not equipped to measure eye movements quantitatively, but we tested qualitatively for deficits in eye control that might account for reaching errors. In no case did the monkey show any signs of nystagmus, and its eyes followed the movement of fruit bits with no obvious impairments. On this basis, it seems unlikely that the visible deficits in reaching produced by PRZ inactivations resulted from problems in eye movement control, but this possibility cannot be excluded.
). A large fraction of the units discharged intensely for several different individual digit movements, which led to the conclusion that the activity was related to stabilizing unused digits in an extended position when an individual digit needed to be flexed. A lesion of the motor cortex did not diminish the intensity of rubral discharge, but individual digit motion was now impaired and discharge occurred preferentially during grouped extension (and in some cases flexion) of all of the fingers. This fits well with the deficit in hand shaping and finger extension at site J7.5b. At the same site, unit discharge showed a complex temporal pattern compatible with the finger extensor activity required in three phases of the tube task (Fig. 6): as the hand was lifted from the touch-pad (but not during the transport phase of the reach), as the hand was being shaped for entry into the tube, and while digits 3-5 were being stabilized in extension to allow the thumb and index fingers to manipulate and grasp a food object. Further anterior in NIA (J5.5), the deficit was probably also one of hand shaping, although it presented as a pinch grasp failure. Evidently, the pinch grasp became ineffective because the animal lost its normal capacity to stabilize the unused digits by flexing them against the palm of the hand. Moving to the next stage of processing, other studies have demonstrated preferential linkages between rubral neurons and the extrinsic extensors and flexors of the fingers (Miller et al. 1993; Sinkjaer et al. 1995). In combination, these various results begin to specify how synergistic extension and flexion of the fingers, and perhaps other synergies that help to shape the hand, may be represented through an interconnected portion of the cerebello-rubro-spinal pathway. If similar analyses were available for other synergies, one could begin to define the distributed synergy maps comprising a muscle-based coordinate system for controlling voluntary movements of the arm (Miller and Houk 1995; Miller et al. 1993).
). Some authors have stressed the coordination of temporal patterns in motor commands so as to achieve smooth movements with accurate endpoints (e.g., Hore et al. 1991), whereas others have stressed the coordination of spatial patterns of muscle activity to achieve synergistic control of multijoint movements (e.g., Thach et al. 1992a). It is likely that both views are correct and that the overall function of the cerebellum is to ensure spatiotemporal coordination in the large population of nuclear cells that participate in any given movement. The microinjections of muscimol used in the present study would be expected to inactivate select foci within this population thus disrupting the spatial aspect of coordination, which is precisely what we observed. When we inactivated anterior foci, we disrupted different hand components of arm coordination, leaving reach components intact. Our interpretation is that we blocked elements of the population that send commands to distal forelimb muscles. In contrast, when we inactivated posterior foci, we disrupted predominantly reach components of forelimb coordination presumably by blocking elements of the population that send commands to proximal arm muscles.
, 1993; van Kan et al. 1993b) to reveal a site specificity as clear as that seen with muscimol inactivation may be a reflection of how the cerebellum normally performs coordination. Discrete regions of the cerebellar nuclei are not activated in isolation; instead, large populations of neurons fire in combination (Fortier et al. 1989; Thach et al. 1993; van Kan et al. 1993b). Although virtually all of the intermediate nuclear cells seem to require a hand component in the task to ensure high rates of discharge (Gibson et al. 1996; van Kan et al. 1994), this does not mean that all of these cells have a hand control function. Instead, it probably reflects some involvement of hand shaping, gripping, and manipulation, along with aiming and stabilization of reach, in those voluntary tasks that engage the intermediate cerebellum. The perhaps highly specific contribution of any given neuron to task performance may be obscured by the relatively large number of cells participating in the movements, unless more detailed correlations with specific skills or specific muscle activity patterns are attempted.
). On the basis of known properties of granule cells (D'Angelo et al. 1995), such mossy fiber inputs would produce asynchronous patterns of activity in PFs rather than the synchronous beams required in the Thach et al. (1992a) model.
). This network is composed of recurrent connections from the intermediate zone of nuclear cells to magnocellular red nucleus and, via a ventral thalamic relay, to primary motor cortex. Collaterals of rubrospinal and corticospinal fibers then innervate precerebellar neurons in the lateral reticular and pontine nuclei, and these sources of mossy fiber input send collaterals to innervate the intermediate nuclear cells to close the positive feedback loop. The buildup of positive feedback in this network appears to be responsible for recruiting the large population of rubral and motor cortical cells that are needed to activate the many muscles in the limb that participate in a voluntary movement (Houk et al. 1993).
). Anteroposterior bands of PCs are innervated by climbing fibers (CFs) that originate from discrete sites within the inferior olivary nucleus (Voogd and Bigaré 1980). These bands of PCs then converge upon small clusters of nuclear cells, the same clusters innervated by collaterals of the CFs (Oscarsson 1980; Voogd and Bigaré 1980). In several models of the cerebellum, these olivocerebellar complexes are considered as adaptive networks for regulating motor commands in the cerebellar nuclei (cf. Houk et al. 1996).
), in a pattern that parallels other mammalian species (Voogd and Bigaré 1980). NIA is innervated by anteroposterior bands of PCs, designated C1 and C3, which receive their CF input from the rostral portion of the dorsal accessory olive (rDAO). In contrast, NIP is innervated by the anteroposterior band of PCs designated C2 which receives its CF input from the rostral portion of the medial accessory olive (rMAO). The organizational features that distinguish rDAO and rMAO have been studied most extensively in the cat. Bidirectional transport of a neuroanatomic tracer (wheat germaglutin-horse radish peroxidase; WGA-HRP) demonstrated that rDAO connections with NIA are organized somatotopically with the hindlimb represented anterior and medially and the forelimb represented posterior and laterally (Gibson et al. 1987), and a fine representation within forelimb NIA also has been demonstrated (Trott et al. 1990). The low-threshold tactile receptive fields of rDAO neurons emphasize the distal limb, whereas the predominantly proprioceptive receptive fields of rMAO include many cells related to shoulder rotation (Gellman et al. 1983, 1985).
).
| |
ACKNOWLEDGEMENTS |
|---|
This work was supported by National Institutes of Health Grants MH-48185 to J. C. Houk and EY-07342 to J. F. Baker.
| |
FOOTNOTES |
|---|
Present address of C. R. Mason: Dept. of Neurosurgery, University of Minnesota, Minneapolis, MN 55455.
Address for reprint requests: J. C. Houk, Dept. of Physiology, Northwestern University Medical School, 303 E. Chicago Ave., Ward 5.315, Chicago, IL 60611-3008.
Received 20 December 1996; accepted in final form 30 September 1997 .
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
New Vistas,
edited by
S. L. Palay,
and V. Chan-Palay
. New York: Springer-Verlag, 1982, p. 440-454 This article has been cited by other articles:
|
|
 |
|
  V. Prevosto, W. Graf, and G. Ugolini Cerebellar Inputs to Intraparietal Cortex Areas LIP and MIP: Functional Frameworks for Adaptive Control of Eye Movements, Reaching, and Arm/Eye/Head Movement Coordination Cereb Cortex, May 22, 2009; (2009) bhp091v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Ilg, M. A. Giese, E. R. Gizewski, B. Schoch, and D. Timmann The influence of focal cerebellar lesions on the control and adaptation of gait Brain, November 1, 2008; 131(11): 2913 - 2927. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Castiello and C. Begliomini The Cortical Control of Visually Guided Grasping Neuroscientist, April 1, 2008; 14(2): 157 - 170. [Abstract] [PDF]
|
|
|
|
 |
|
 D. S. Soteropoulos and S. N. Baker Bilateral representation in the deep cerebellar nuclei J. Physiol., February 15, 2008; 586(4): 1117 - 1136. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Pizzimenti, W. G. Darling, D. L. Rotella, D. W. McNeal, J. L. Herrick, J. Ge, K. S. Stilwell-Morecraft, and R. J. Morecraft Measurement of Reaching Kinematics and Prehensile Dexterity in Nonhuman Primates J Neurophysiol, August 1, 2007; 98(2): 1015 - 1029. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Mason, C. M. Hendrix, and T. J. Ebner Purkinje Cells Signal Hand Shape and Grasp Force During Reach-to-Grasp in the Monkey J Neurophysiol, January 1, 2006; 95(1): 144 - 158. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Konczak, B. Schoch, A. Dimitrova, E. Gizewski, and D. Timmann Functional recovery of children and adolescents after cerebellar tumour resection Brain, June 1, 2005; 128(6): 1428 - 1441. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Shipp The importance of being agranular: a comparative account of visual and motor cortex Phil Trans R Soc B, April 29, 2005; 360(1456): 797 - 814. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. N. Holdefer, J. C. Houk, and L. E. Miller Movement-Related Discharge in the Cerebellar Nuclei Persists After Local Injections of GABAA Antagonists J Neurophysiol, January 1, 2005; 93(1): 35 - 43. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Monzee and A. M. Smith Responses of Cerebellar Interpositus Neurons to Predictable Perturbations Applied to an Object Held in a Precision Grip J Neurophysiol, March 1, 2004; 91(3): 1230 - 1239. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Monzee, T. Drew, and A. M. Smith Effects of Muscimol Inactivation of the Cerebellar Nuclei on Precision Grip J Neurophysiol, March 1, 2004; 91(3): 1240 - 1249. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. P. Goodkin and W. T. Thach Cerebellar Control of Constrained and Unconstrained Movements. I. Nuclear Inactivation J Neurophysiol, February 1, 2003; 89(2): 884 - 895. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Pong, K. M. Horn, and A. R. Gibson Spinal Projections of the Cat Parvicellular Red Nucleus J Neurophysiol, January 1, 2002; 87(1): 453 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Horn, M. Pong, S. R. Batni, S. M. Levy, and A. R. Gibson Functional Specialization Within the Cat Red Nucleus J Neurophysiol, January 1, 2002; 87(1): 469 - 477. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. L. E. Van Kan and M. L. McCurdy Role of Primate Magnocellular Red Nucleus Neurons in Controlling Hand Preshaping During Reaching to Grasp J Neurophysiol, April 1, 2001; 85(4): 1461 - 1478. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gruart, G. Guillazo-Blanch, R. Fernandez-Mas, L. Jimenez-Diaz, and J. M. Delgado-Garcia Cerebellar Posterior Interpositus Nucleus as an Enhancer of Classically Conditioned Eyelid Responses in Alert Cats J Neurophysiol, November 1, 2000; 84(5): 2680 - 2690. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. R. Robinson Role of the Cerebellar Posterior Interpositus Nucleus in Saccades I. Effect of Temporary Lesions J Neurophysiol, September 1, 2000; 84(3): 1289 - 1302. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. H. Martin, S. E. Cooper, A. Hacking, and C. Ghez Differential Effects of Deep Cerebellar Nuclei Inactivation on Reaching and Adaptive Control J Neurophysiol, April 1, 2000; 83(4): 1886 - 1899. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Timmann, S. Watts, and J. Hore Failure of Cerebellar Patients to Time Finger Opening Precisely Causes Ball High-Low Inaccuracy in Overarm Throws J Neurophysiol, July 1, 1999; 82(1): 103 - 114. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Hoover and P. L. Strick The Organization of Cerebellar and Basal Ganglia Outputs to Primary Motor Cortex as Revealed by Retrograde Transneuronal Transport of Herpes Simplex Virus Type 1 J. Neurosci., February 15, 1999; 19(4): 1446 - 1463. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
